模糊因果聚类模型在高炉焦比预测中的应用

Application of Fuzzy Causal Classification Model for Forecasting Coke Ratio in Blast Furnace

通过通径分析,对高炉现场采集的数据进行处理,在给定的描述高炉系统的诸多变量中,利用最小剩余通径系数确定影响目标函数的主要变量因素·将诸因素关系处理为直接通径和间接通径,并对其进行了排序,找出了影响指定目标函数:焦比的主要直接通径和间接通径·综合直接通径和间接通径效果,确定了高炉炉顶温度、料批、矿批重、焦炭负荷和[Si]既是影响焦比的直接原因,也是其他因素对焦比作用的间接原因·

By means of path analysis, the data collected in situ were processed. Among the variables given to describe blast furnace system, the main variable influencing factors on objective function were determined according to the minimum remaining path coefficient. The relation among those factors was simplified and classified into two groups, i.e., direct paths and indirect paths, then they were ranked to find out the influenced objective functions, i.e., the direct and indirect paths to coke ratio. Analyzing comprehensively the effects of the two paths, a conclusion is drawn that the furnace roof temperature, batch weight, ore weight ,coke burden and [Si] are not only the direct influencing factors, but also the indirect influencing factors on coke ratio.